Sei sulla pagina 1di 7

# Pet.Sci.

(2009)6:91-97

91

DOI 10.1007/s12182-009-0017-9

## Numerical simulation of predicting and reducing

solid particle erosion of solid-liquid two-phase flow
in a choke
Li Guomei1, 2, Wang Yueshe1*, He Renyang2, Cao Xuewen3, Lin Changzhi4
and Meng Tao2
State Key Laboratory of Multiphase Flow in Power Engineering, Xian Jiaotong University, Xian, Shaanxi 710049, China
China Special Equipment Inspection and Research Institute, Beijing 100013, China
3
School of Transport & Storage and Civil Engineering, China University of Petroleum, Qingdao, Shandong 266555, China
4
Petroleum Exploration and Production Research Institute, Sinopec, Beijing 100083, China
1
2

Abstract: Chokes are one of the most important components of downhole flow-control equipment. The
particle erosion mathematical model, which considers particle-particle interaction, was established and
used to simulate solid particle movement as well as particle erosion characteristics of the solid-liquid
two-phase flow in a choke. The corresponding erosion reduction approach by setting ribs on the inner
wall of the choke was advanced. This mathematical model includes three parts: the flow field simulation
of the continuous carrier fluid by an Eulerian approach, the particle interaction simulation using the
discrete particle hard sphere model by a Lagrangian approach and calculation of erosion rate using semiempirical correlations. The results show that particles accumulated in a narrow region from inlet to outlet
of the choke and the dominating factor affecting particle motion is the fluid drag force. As a result, the
optimization of rib geometrical parameters indicates that good anti-erosion performance can be achieved
by four ribs, each of them with a height (H) of 3 mm and a width (B) of 5 mm equaling the interval
between ribs (L).

Key words: Solid-liquid two-phase flow, discrete particle hard sphere model, choke, erosion rate, antierosion, numerical simulation

1 Introduction
The mechanical damage to the surface caused by the
impact of solid particles has been a serious problem in a
variety of engineering applications. Any industrial process
involving the transportation of solid particles entrained in
the fluid phase can be subject to erosion damage (Humphrey,
1990; Finnie, 1995). In oil and gas production, the solid
particles, which were used as proppants and carried by
fracturing fluids with a high velocity in sand fracturing
operations, can cause serious damage to downhole flowcontrol equipment (chokes) as well as the surface of casing
walls (Economides and Nolte, 2002; McCasland et al,
2004; Vincent et al, 2004). During oil-field water injection
operations, particles entrained in continuous fluids can also
cause damage to the casing wall (Jordan, 1998; Richardson
et al, 1986). Depending on the actual conditions, the erosion
damage may be severe and extremely expensive, as it may
be frequently necessary to replace or repair the device or
component that is exposed and susceptible to the erosive
environment (McCasland et al, 2004; Jordan 1998; McLaury
* Corresponding author. email: wangys@mail.xjtu.edu.cn

## et al, 1996; Wallace et al, 2004). Hence, it becomes more

and more important to predict the erosion caused by particle
impact accurately and to develop corresponding erosion
reduction methods.
Computational Fluid Dynamics (CFD) has been used
in the research on solid particle erosion for many years.
CFD-based erosion prediction process includes several
different aspects (Zhang, 2006): flow modeling, particlefluid interaction, particle-particle interaction, particle-wall
interaction and particle erosion modeling. Each aspect by
itself is very complex, and many researchers have made
great efforts in order to better understand the mechanisms.
The current erosion computational models are established
based on different mechanisms, which show satisfactory
application to predicting particle movement characteristics,
calculating the erosion rate of wall surfaces and improving
particle tracking in order to reduce erosion (Chen et al, 2006;
Fan et al, 2004; Forder et al, 1998; Habib and Badr, 2004;
McCasland et al, 2004; McLaury and Wang 1997; Song et
al, 1996; Yao et al, 2002). However, most of these models
neglect the influence of particle movement on the fluid as well
as particle-particle interactions. These are one-way coupling
methods and only applicable to the conditions of low volume
fraction of the discrete particle phase. The four-way coupling

Pet.Sci.(2009)6:91-97

92

## computational model (Elghobashi, 1994), which takes

particle-fluid interaction and particle-particle interaction into
consideration at the same time, has been rarely reported in the
literature, especially in the study of erosion of well pipeline
systems.
The purpose of this paper is to provide deeper
understanding of solid particle erosion characteristics in
the choke and to change the flow field geometry in order to
reduce erosion. In this paper, the discrete particle hard sphere
model was used to simulate particle-particle interaction;
after obtaining information on particle movement (impact
velocity and impact angle), the semi-empirical correlations
were used to calculate particle erosion rate. Fortunately,
this model not only takes particle-fluid interaction into
consideration but particle-particle interaction as well, and is
a four-way coupling method. Therefore, this model avoids
the shortcomings of former models mentioned above and is
applicable to the case of high volume fractions of the particle
phase.

2 Mathematical model

## 2.2 Discrete particle hard sphere model

2.2.1 Inter-particle collision model

## Inter-particle collision is described by a hard sphere

model. The hard sphere model is based on binary quasiinstantaneous collisions. It neglects particle deformation
during collision, resulting in a high calculation efficiency
(Crowe et al, 1998).
Collision of two particles can be shown in Fig. 1. All the
post-collision velocities can be expressed as follows:
(5)
(6)
(7)
(8)
for the case where the two spheres slide, and

## The mathematical model includes the following three

parts, namely the continuous carrier fluid flow field simulation
by an Eulerian approach, particle-particle interaction
simulation using a discrete particle hard sphere model by a
Lagrangian approach, and erosion calculation using semiempirical correlations.

(9)
(10)

(11)

## The equations of continuous flow are derived from the

volume averaged Navier-Stokes equations, which take into
account the influence of fluid volume fraction and drag force
between the fluid and particle phases. The continuity and
momentum equations are expressed as follows:
(1)

(2)

(12)
for the case where two spheres stop sliding during the
collision process,where superscript 0 means before collision;
s is fraction coefficient; e is the coefficient of restitution; n
is the unit normal vector from particle 1 to particle 2 at the
moment of contact; m is the particle mass; R is particle radius;
v is particle velocity; is angular velocity; G0 is the relative
velocities between particle centers before collision, and Gct0 is
the tangential component of the relative velocity; t is the unit
tangential vector at contact point.
0
1

## where , and p are the fluid density, velocity, and pressure,

respectively; g is gravity acceleration; is the fluid shear
stress; is the volume fraction of fluid; and fdrag is volumetric
fluid-particle interaction force, which can be given as:

1
0
2
1

(3)

(4)
where V and Vpi are the volume of a computational cell and
the volume of particle i inside this cell, respectively; Fdrag is
the fluid drag force for an individual particle; n is the number
of particles in the cell.
The flow is turbulent, and the fluid turbulence is treated
with the standard
turbulent model.

n
2

## The coupling action between fluid and particle can be

expressed through the fluid drag force Fdrag of an individual
particle. Using the modified fluid drag force correlation (Di
Felice, 1994), the fluid drag force can be described as follows:

Pet.Sci.(2009)6:91-97

93

(13)
with

## where dp is the particle diameter; uf is the fluid velocity; up is

the particle velocity; Cd0 and Rep are the fluid drag coefficient
and particle Reynolds number, respectively, and they can be
expressed as follows:

## 2.4 Criterion for occurrence of collision of two

particles in the hard sphere model
In the hard sphere model system, the initiative search
approach is used to judge the binary collisions. It is supposed
that the two particles i and j happen to collide after the time
tc, so Rij=rij+vijtc, as a result tc can be calculated with the
following expression (Kang and Guo, 2006):
(9)
with

## where r is the location vector of particles.

If tc<0 or if rij.vij>0, the two particles will not collide.
2.2.3 Equation of particle motion

## The behavior of all particles is simulated by a Lagrangian

approach. Because inter-particle collision is described by
the hard sphere model, the inter-particle force and particle
shape deformation can be neglected (Crowe et al, 1998).
Thus forces acting on an individual particle mainly include
gravitational and fluid drag forces during particle movement
(Zhang, 2006). Other forces such as virtual mass force,
pressure gradient force, Saffman lift force, Magnus lift
force and Basset history force acting on a particle could
be neglected in this study (Habib and Badr 2004; Habib et
al, 2007; Meng and Van der Geld 1991). The motion of an
individual particle is determined by Newtons second law of
motion. The equation of particle can be written as follows:
(14)
where mp is the particle mass.

## 2.3 Semi-empirical correlations used to calculate

particle erosion rate
According to the dynamic characteristics of solid-liquid
two-phase flow (which is different from gas-solid twophase flow) and the material characteristics of the choke,
the calculation model proposed by Mengtrk and Sverdrup
(1979) was applied in this study:

## 3 Physical problem and simulation

conditions
The computational domain is a two-dimensional
rectangular region of 0.1m0.038m, which is the same as the
actual size of the choke, the inlet radius is 0.024 m and the
outlet diameter is 0.018 m, as shown in Fig. 2. The hatching
boundary is a wall boundary condition, where a no-slip
boundary condition is imposed for fluid flow. In the particlewall collision, the interaction of a particle with the wall is
modeled in the same manner as a particle-particle collision,
and the wall is considered as one particle with infinitely large
diameter (Crowe et al, 1998). The top dashed axis boundary
is a symmetry boundary and zero normal gradients are used
for all variables. As for particles, the reflective boundary
condition is used on the axis boundary. The direction of
gravity goes along the X direction. The fluid flows into
the choke from the inlet, where a velocity inlet boundary
condition is imposed for continuous flow, and the fluid flows
out through the outlet, where a pressure outlet boundary
condition is imposed.
Axis
Inlet

Inner space of
choke

Outlet

Wall

## Fig. 2 Computational domain and coordinate system

(15)

(16)
where Ev is the particle erosion rate, mm3/g; W1 is the wall
impact velocity of the solid particle; and 1 is the impact
angle of the solid particle.

## The continuous fluid is water with a density of 1,000

kg/m 3 and a dynamic viscosity of 1.00510 -3 Pa . s. The
calculation temperature is room temperature, and the flow in
the choke is turbulent. The velocity at inlet section follows
the 1/7 power function distribution, and the turbulent center
velocity was chosen as 5 m/s.
The particles all have the same diameter (0.4 mm)
and density (2,650 kg/m 3). The friction coefficient and
restitution coefficient are 0.4 and 0.9, respectively. In the

Pet.Sci.(2009)6:91-97

96
1.0

1.0

1.0
A

O
0.5

0.5

Probability, %

1.0

0.5

1.0

0.5

0.5

1.0

0.5

1.0

D
0.5

0
1.0

1.0

0.5

H
0.5

Velocity, m/s
Fig. 7 Probability distribution of particle-wall impact velocity

## Therefore setting ribs on the inner wall of the choke can

significantly reduce particle average impact velocity, which is
a major reason for reducing erosion by setting ribs.
In order to discuss the erosion severity of the choke with
different rib geometrical parameters, the relative erosion rate
E/E0 is defined, namely the ratio of erosion rate of choke
wall after setting ribs E and that of before setting ribs E0. The
smaller the relative erosion rate is, the better the anti-erosion
performance. And the relative erosion rates at different rib
geometrical parameters are shown in Fig. 8. The mean relative
erosion rate of the choke is from 0.04 to 0.61, with an average
value of 0.14, after setting ribs. That is to say setting ribs can
improve the anti-erosion performance of the choke greatly,
which show good agreement with the above conclusions.
Fig. 8 also indicates that rib type F has the smallest relative
rate, in other words, the best anti-erosion performance can
be achieved when there are 4 ribs in the choke, each of them
with a height (H) of 3 mm and a width (B) of 5 mm spaced 5
mm apart (L).

1.0

1.0

0.8

0.4
0.28
0.2

0.138

0.14
0.036

0.024

0.028

## This work is supported by the Fund of Innovation

Research Group of National Natural Science Foundation of
China (Grant NO. 5052160450323001) and Major Program
of National Natural Science Foundation of China (Grant No.
50536020).

0.008
0.004

0.0
O

## The following conclusions can be drawn from the above

discussion:
1) The discrete particle hard sphere model can simulate
the dynamic process of solid particle movements and the
non-uniform discrete characteristics of particle distribution
in a choke. At the same time, this model can provide detailed
information on the impact velocity, impact angle and impact
location of an individual particle.
2) Particles accumulate densely in a narrow domain from
inlet to outlet of the choke, where the largest particle velocity
occurs. The dominating factor affecting particle motion in
solid-liquid two-phase flow is the fluid drag force, but particle
collisions also play an important role.
3) When the choke has no ribs, only 40% of particles have
a velocity of less than 0.5 m/s. With ribs in the choke, over
90% of particles have a velocity of less than 0.5 m/s. That is
to say the average impact velocity of particles lies in the low
impact velocity domain after setting ribs.
4) Using ribs can greatly improve the anti-erosion
performance of the choke, and the relative best anti-erosion
performance can be achieved when setting 4 ribs, each of
them with a height (H) of 3 mm and a width (B) of 5 mm at 5
mm spacings (L).

Acknowledgements

0.61

0.6

5 Conclusions

Mean

Type of ribs
Fig. 8 Relative erosion rate at different rib geometrical parameters

Reference
Chen X H, McLaury B, S and Shirazi S A. Numerical and experimental
investigation of the relative erosion severity between plugged tees

Pet.Sci.(2009)6:91-97

and elbows in dilute gas/solid two-phase flow. Wear. 2006. 261 (7-8):
715-729
Crowe C T, Sommerfeld M and Tsuji Y T. Multiphase Flows with
Droplets and Particles. Boca Raton: CRC Press. 1998. 112-165
Di Felice R. The voidage function for fluid-particle interaction systems.
International Journal of Multiphase Flow. 1994. 20(1): 153-159
Economides M J and Nolte K G. Reservoir Stimulation (3rd Edition).
Translated by Zhang B P. Beijing: Petroleum Industry Press. 2002.
133-158
Elghobashi S. On predicting particle-laden turbulent flows. Applied
Scientific Research. 1994. 52(2): 309-329
Fan J R, Luo K, Zhang X Y, et al. Large eddy simulation of the antierosion characteristics of the ribbed-bend in gas-solid flows. Journal
of Engineering for Gas Turbines and Power. 2004. 126(3): 672-679
Finnie I. Some reflections on the past and future of erosion. Wear. 1995.
186-187(Part 1): 1-10
Forder A, Thew M and Harrison D. A numerical investigation of solid
particle erosion experienced within oilfield control valves. Wear.
1998. 216(2): 184-193
Habib M A and Badr H M. Numerical calculations of erosion in an
abrupt pipe contraction of different contraction ratios. International
Journal of Numerical Method in Fluids. 2004. 46(1): 19-35
Habib M A, Badr H M, Mansour R B, et al. Erosion rate correlations of
a pipe protruded in an abrupt pipe contraction. International Journal
of Impact Engineering. 2007. 34(8): 1350-1369
Humphrey J A C. Fundamentals of fluid motion in erosion by solid
particle impact. International Journal of Heat and Fluid Flow. 1990.
11(3): 170-195
Jordan K. Erosion in multiphase production of oil & gas. Corrosion.
1998. Paper No. 058. NACE. 1998
Kang L Q and Guo L J. Eulerian-Lagrangian simulation of aeolian sand
transport. Powder Technology, 2006. 162(2): 111-120
Lindsley B A and Marder A R. The effect of velocity on the solid particle
erosion rate of alloys. Wear. 1999. 225-229(Part 1): 510-516
McCasland M, Barrilleaux M, Gai H, et al. Predicting and mitigating

97

## erosion of downhole flow control equipment in water-injector

completions. SPE Annual Technical Conference and Exhibition,
26-29 September 2004, Houston, Texas (SPE 90179)
McLaury B S, Shirazi S A, Shadley J R, et al. Modeling erosion in
chokes. ASME Fluids Engineering Division. 1996. 236(1): 773-781
McLaury B S and Wang J. Solid particle erosion in long radius
elbows and straight pipes. SPE Annual Technical Conference and
Exhibition, 5-8 October 1997, San Antonio, Texas (SPE 38842)
Meng H and Van der Geld C W M. Particle trajectory computations
in steady non-uniform liquid flows. ASME Fluids Engineering
Division. 1991. 118: 183-93
Mengtrk M and Sverdrup E F. Calculated tolerance of a large electric
utility gas turbine to erosion damage by coal gas ash particles. In:
Erosion: Prevention and Useful Application. Edited by Adler W F.
Philadelphia: ASTM Spec. Tech. Publ. 1979. 193-2224
Richardson J W, Le K M and Riggs K R. Production casing erosion
through annular injection. SPE Annual Technical Conference and
Exhibition, 2-5 October 1988, Houston, Texas (SPE 18056)
Song X Q, Lin J Z, Zhao J F, et al. Research on reducing erosion by
adding ribs on the wall in particulate two-phase flows. Wear. 1996.
193(1): 1-7
Vincent M C, Miller H B, Milton-Tayler D, et al. Erosion by proppant:
a comparison of the erosivity of sand and ceramic proppants during
slurry injection and flowback of proppant. SPE Annual Technical
Conference and Exhibition, 26-29 September 2004, Houston, Texas
(SPE 90604)
Wallace M S, Dempster W M, Scanlon T, et al. Prediction of impact
erosion in valve geometries. Wear. 2004. 256(9-10): 927-936
Yao J, Fan J R and Cen K C. Numerical investigation of a new method
for reducing bends erosion from particles impacts. Chinese Journal
of Chemical Engineering. 2002. 10 (1): 93-101
Zhang Y L. Application and Improvement of Computational Fluid
Dynamics (CFD) in Solid Particle Erosion Modeling. Ph.D Thesis.
The University of Tulsa. 2006
(Edited by Sun Yanhua)